{
  "id": 212635,
  "title": "Imbalanced techniques",
  "url": "/competitions/siim-isic-melanoma-classification/discussion/212635",
  "author_name": "",
  "post_date": "2021-01-19T16:45:45.103000",
  "votes": 3,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Hi everyone,</p>\n<p><a href=\"https://challenge2019.isic-archive.com/\" target=\"_blank\">https://challenge2019.isic-archive.com/</a><br>\nI used the dataset from this competition.This was highly imbalanced dataset .<br>\nFor this i used </p>\n<ul>\n<li>Weighted classes approach<br>\nbut no use , the prediction biasing towards to only single class.<br>\nSo i am planning to use the upsampling by using the data augmentation only for the minority classes.<br>\nBut i didn't get an idea how to apply augmetation only for minoirty classes and how to create a generator from this .<br>\nor </li>\n<li>IS there any other way to solve the imbalancing.</li>\n</ul>\n<p>Sorry for my english.</p>",
  "messages": [
    {
      "id": 1160071,
      "postDate": "2021-01-19T16:45:45.103Z",
      "content": "<p>Hi everyone,</p>\n<p><a href=\"https://challenge2019.isic-archive.com/\" target=\"_blank\">https://challenge2019.isic-archive.com/</a><br>\nI used the dataset from this competition.This was highly imbalanced dataset .<br>\nFor this i used </p>\n<ul>\n<li>Weighted classes approach<br>\nbut no use , the prediction biasing towards to only single class.<br>\nSo i am planning to use the upsampling by using the data augmentation only for the minority classes.<br>\nBut i didn't get an idea how to apply augmetation only for minoirty classes and how to create a generator from this .<br>\nor </li>\n<li>IS there any other way to solve the imbalancing.</li>\n</ul>\n<p>Sorry for my english.</p>",
      "rawMarkdown": "Hi everyone,\n\nhttps://challenge2019.isic-archive.com/\nI used the dataset from this competition.This was highly imbalanced dataset .\nFor this i used \n- Weighted classes approach\nbut no use , the prediction biasing towards to only single class.\nSo i am planning to use the upsampling by using the data augmentation only for the minority classes.\nBut i didn't get an idea how to apply augmetation only for minoirty classes and how to create a generator from this .\nor \n- IS there any other way to solve the imbalancing.\n\nSorry for my english.",
      "votes": 3
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "1160071": "Hi everyone,\n\nhttps://challenge2019.isic-archive.com/\nI used the dataset from this competition.This was highly imbalanced dataset .\nFor this i used \n- Weighted classes approach\nbut no use , the prediction biasing towards to only single class.\nSo i am planning to use the upsampling by using the data augmentation only for the minority classes.\nBut i didn't get an idea how to apply augmetation only for minoirty classes and how to create a generator from this .\nor \n- IS there any other way to solve the imbalancing.\n\nSorry for my english."
  }
}